Retweeting for @rabihmgeha, @DxRxEdu, the @CPSolvers team, and the VMR community, a thread about how and when diagnostic triads might lead to diagnostic errors:
1/13

Diagnostic triads are memory aids that list 3 findings characteristic of a single diagnosis. (Siblings tetrads and pentads list 4 and 5 findings, respectively.) When we see those 3 findings together, we are taught to think of that one condition.
2/13

Implicitly, many of us learning these triads are taught to believe two 'scoring rules': that all 3 findings must occur for the disorder to be present; and, that without all 3 findings, the disorder is disproved. How could this lead to diagnostic error?
3/13

First, not all patients with a specific disorder will have every single finding 100% of the time. For example, only 6 - 15% of patients with proved acute aortic dissection will exhibit asymmetry of arm pulses or blood pressures. JAMA 2002; 287: 2262 - 2272.
4/13

The proportion of patients with the disease that have a finding can be understood as the sensitivity of the 'test', i.e. the presence of the finding is a 'positive' result. ACP J Club 2002 Jul/Aug; A11 - A12.
5/13

If we use the absence of arm pulse asymmetry or blood pressure to exclude aortic dissection, we will mistakenly discard this diagnosis in 85 to 94% of patients with proven dissection. JAMA 2000; 284: 869 - 875.
6/13

Second, as a corollary, not all patients with a disease will have all 3 findings in a diagnostic triad 100% of the time. For example, half of patients with pericardial tamponade were found to have none of 3 findings: hypotension, elevated JVP, and muffled heart sounds.
7/13

Thus, if we use the absence of the 3 findings of Beck's triad to exclude pericardial tamponade, we will mistakenly discard this diagnosis in 50% of patients: World J Emerg Med 2017; 8: 29 - 33.
8/13

Third, very few clinical findings are 100% specific for a single disorder; often there are many causes of false positives. For instance, sinus tachycardia could be from depletion of intravascular volume, yet there are many other causes of this finding.
9/13

If we interpret the presence of sinus tachycardia as proving volume depletion, we would close prematurely on this diagnosis without safely considering some important alternatives.
10/13

Fourth, very few triads of clinical findings are 100% specific. In a patient with all 3 triad findings, if we interpret this to mean the disorder is certain, we would close prematurely on this diagnosis and could miss the correct diagnosis among the alternatives.
11/13

If used without care, diagnostic triads, tetrads, and pentads can contribute to diagnostic error in any of these 4 ways. Some have advocated that we abandon their use: Med Education 2019; 53: 110 - 112.
12/13

If we want to continue using diagnostic triads, we should also teach the likelihood ratios or sensitivity/specificity of each individual finding, as well as the likelihood ratios or sensitivity/specificity of all 3 findings together.
13/13